Re: slosh (was Re: Armadillo May 2, 2004)

At 07:08 PM 5/3/2004 -0400, Henry wrote:
>On Mon, 3 May 2004, John Carmack wrote:
> > If we find slosh to be an issue, our easiest correction will be to float a
> > bunch of stainless steel balls in the tank, which is effectively what they
> > did on Vanguard.
>
>Mmm, I'm not sure that will be very effective.  (I assume by "balls" you
>mean "floats".)  For cylindrical tanks, the main slosh mode is the surface
>going up on one side and down on the other, with actual liquid motion
>mostly somewhat below the surface.  I don't think floats are going to damp
>this very well.

I just went back and checked the reference that I had read that in, and it 
was Jupiter, not Vanguard:

Nasa Technical Paper 3653, P20

Liquid propellant sloshing became a problem due to the loss of the second 
Jupiter vehicle.
Analytical and experimental data and approaches were not available to 
characterize the problem. The
first attempt was to place the full-size propellant tank on a railroad car 
then bump it against the rail end
stop as an excitation mechanism. In terms of today's standards, this was 
not a well-controlled experiment.
Large perforated cylinders with sealed spheres (commode floats) were 
floated in the tank and
again tested, showing their good damping characteristics. The next Jupiter 
was flown successfully using
this system. New technologies resulted as the program moved forward:
1. Analytical representation of fluid dynamic characteristics
2. Equivalent mechanical analog of the complex analytical equations 
(pendulum or mass spring
slosh model)
3. Scale model testing techniques, instrumentation, and data evaluation
4. Development of slosh baffles that became an integral part of the 
structural stiffening ring,
saving weight over the floating (beer) cans.